BULLETIN OF THE FLORIDA STATE MUSEUM BIOLOGICAL SCIENCES Volume 10 Number 7 VARIATION AMONG THE SOUTHEASTERN CROWNED SNAKES, GENUS TANTILLA Sam Rountree Telford, Jr. ·t Of . -. /853 ' UNIVERSITY OF FLORIDA Gainesville 1966 Numbers of the BUL]bETIN OF THE FLORIDA STATE MUSEUM are pub- lished at irregular intervals. Volumes contain about 300 pages} and are not nec- essarily completed in any one calendar year. WALTER AUF'FENBERG, Managing Editor OLIVER L. AUSTIN, JR., Editor Consultants. for this issue: WALTER AUFFENBERG HOBART M. SMITH Communications concerning purchase or exchange of the publicati6n and all manuscripts should be_addressed to the Managing Editor,of the Bullatin, Florida State Museum, Seagle Building, GainesVille, F16rida. 82601 Published May 25, 1966 Price for this issue $.65 VARIATION AMONG THE SOUTHEASTERN CROWNED SNAKES , GENUS TANTILLA SAM ROUNTREE TELFORD, JR.1 SYNoPSIS: The coronata group of crowned snakes in southeastern United States includes three species: Tantilla coronata Baird and Girard, 1853, and two spe- cies herein described, T. retitta and T. oolitica. T. coronata ranges through- out southeastern United States, but does not enter peninsular Florida south of the Suwannee River; T. oolitica is restricted to Dade and Monroe counties in southeastern Florida; T. relicta inhabits peninsular Florida. Variation in characters of pattern, scutellation, body proporti6ns, and habitat support recognition of three subspecies in. T. relicta: T. r. neitti occurs in sand- hills and mesic hammocks of northcentral Florida; T. r. reticta of central Florida is restricted to scrub habitat where syntopic with neim, but also occurs in sand- hills where allopatric in southcentral Florida; T. r. pamtica inhabits the narrow strip of coastal dunes and scrub in southeastern Florida. The little evidence of gene flow among subspecies is suggestive of secondary intergradation. Two named subspecies 6f T. coronata, wagneri (Jan) and mitrifer Schwartz, are synonymized with T. coronata. The three southeastern species of Tantilla probably were derived from a com- mon precursor of the early Pleistocene, diverging during the successive fluctua- tions of sea level which characterized the glacial-interglacial stages. i The author completed most of this study while a graduate student at the University of Florida, Gainesville and the University of California, Los Angeles, and presented preliminary conclusions at the annual meetings of the American Society of Ichthyologists and Herpetologists at Austin, Texas, in April 1961. He is currently National Institutes of Health Post-doctoral Fellow in the Department of Parasitology, The Institute for Infectious Diseases, University 6f Tokyo, Japan. Manuscript received 21 June 1965.-ED. Telford, Sam Rountree, Jr. 1966. Variation among the southeastern crowned snakes, genus Tantilla, Bull. Florida State Mus., vol. 10, no. 7, pp. 261-304. 262 BULLETIN FLORIDA STATE MUSEUM Vol. 10 TABLE OF CONTENTS Introduction 262 T. oolitica 281 Acknowledgements and T. coronata 285 Abbreviations 268 Gross Ecology 291 Materials and Methods 264 Size and Sexual Maturity 294 Taxonomy 264 Discussion 295 Tantilld felicia 270 Summary 801 T. r. Telicta 278 Key to southeastern Tantilla __-- 802 T. r. neilli 279 Literature cited 308 T. r. pamtica 280 INTRODUCDON Bairdand Girard (1858: 181) described Tantilla coronata from a specimen collected in Kemper. County, Mississippi. The holotype, United States National Museum 1876, serves also as the type of the genus Tantilla, which Baird and Girard defined at the same time, In 1862 Jan described Homalocranion wagneri from a specimen col- lected in "Florida" by a Professor Wagner. Jan and Sordelli (1866: livr. 15, pl. 2, fig. 8) figure what is apparently the type specimen of H. wagneri. Carman (1888: 163) treated H. wagneri as Tantilla coro- nata var. wagneri. Boulenger (1896: 218) referred H. wagneri to the synonymy of Homalocranium coronatum, adopting the generic designation of Dum6ril and Bibron (1854). Cope (1900: 1111) men- tioned briefly that H. wagneri had not been observed by American naturalists, and since the holotype of wagneri had an entire anal plate, it could not belong to the genus Tantilla, in which the anal is divided. In the same work (1900: 1114), he listed H. wagneri as a synonym of Tantilla coronata. Until Blanchard's review (1988) of the United States members of the genus, the name wagneri was con- sidered a synonym of T. coronata. Blanchard resurrected wagneri for peninsular Florida Tantilla, justifying subspecific relationship to T. coronata on grounds of overlap in scale characters. Most workers since 1938 have followed Blanchard's arrangement. Wright and Wright (1957: 788) commented on the polymorphism of Florida speci- mens, in which all possible variations of banded and non-banded head patterns occur in no proved correlation, and speculated that perhaps there is but one highly variable form of Tantilla in southeastern 1966 VARIATION AMONG SOUTHEASTERN SNAKES 268 United States. In the meantime Schwartz (1958) designated montane populations of Tantilla in Tennessee and North and South Carolina as T. coronata mitrifer. In 1959 I was impressed by the fact that all Tantilla I collected from rosemary scrub habitat (Carr, 1940; Laessle, 1958; Telford, 1959) in southcentral Florida had a prominent light band acrOss the parietal area, separating the black head cap from the black nuchal collar, while all that I collected from sandhills (Laessle, 1958) or mesic ham- mock (Carr, 1940) habitats in northeentral Florida had a predom- inantly black head and nuchal collar with no distinct light parietal band. This prompted me to review relationships of southeastern Tantilla in the light of additional material accumulated since Blanch- ard's 1988 review. In the last 5 years I have examined 215 Tantilla coronata from sites throughout the range exclusive of peninsular Florida, and 166 specimens from the peninsula, of populations pre- viously called Tantilla coronata wagneri. Useful taxonomic characters in scutellation are restricted to ven- trals and subcaudals: Examination of 88 maxillae did not prove particularly fruitful in clarifying the status of the populations. But by comparing structure of the hemipenis with pattern, size, .head shape, scutellation, body proportions, and ecology, I am able to in- terpret the polymorphism of peninsular populations. ACKNOWLEGEMENTS AND ABBREVIATIONS For the loan of specimens I should like to thank: W. J. Riemer, University of Florida Collections (UF); B. Martof, University of Georgia (UG); E. M. Burton, Charleston Museum (CHM); D. M. Cochran, United States National Museum (USNM); C. M. Bogert, American Museum of Natural History (AMNH); E. Williams, Museum of Comparative Zoology (MCZ); N. D. Richmond, Carnegie Museum (CM); R. F. Inger and H. Marx, Chicago Natural History Museum (CNHM); the late N. Hartweg, University of Michigan Museum of Zoology (UMMZ); H. M. Smith, University of Illinois (UI); W. E. Duellman, University of Kansas (UK); H. Dundee, Tulane University (TU); and O. T. Owre, University of Miami (UM); K. L. Hansen, Stetson University; C. D. Wilder, Memphis State University; E. L. Modder, Frostproof High School, Frostproofs Florida; and R. L. Pyke and B. Mansell, Jacksonville, Florida. For the contributions of specimens I would like to thank D. R. Paulson, R. H. Mount, W. T. Neill, and A. Beckenbach. The drawings were prepared by M. Fukushima and E. C. Bovee, to whom I am most grateful. Finally, 264 BULLETIN FLORIDA STATE MUSEUM Vol. 10 I should like to thank my colleagues F. W. King, D. A. Rossman, C. W. .Myers, i J. McCrone, A. Arata, H. W. Campbell, M. J. Fou- quette, Jr., W. T. Neill, and especially R. H. Mount for many stimu- lating discussions and much useful advice. MATERIALS AND METHODS All scale counts were made under a dissecting microscope as follows: ventrals as proposed by Dowling (1951); subcaudals from the tail tip to the vent, excluding the terminal scale; dorsals from the left diagonally across the body at the level of the first ventral; at one head-length posterior to the head; at approximately midbody; and at one head-length anterior to the vent. Head patterns were recorded by shading the appropriate portions of a dorsal and left lateral outline drawing of a Tantilla head. Color descriptions of living specimens were based upon Maerz and Paul (1950). All specimens were meas- ured to the nearest half millimeter with a clear plastic ruler. Left maxillae were removed, cleaned gently with a needle, and dried under a light bulb before examination. In every case where extruded hemipenes were not in evidence, sex was determined by examining the reproductive tract. Sexual maturity of males was established on the basis of enlarged, contorted, or turgid vasa deferentia in lafger specimens, and by positive sperm smears in the case of smaller in- dividuals. The presence of ovarian follicles of varying size or of oviductal eggs was considered indicative of sexual maturity in fe- males, in contrast to uniformly small follicles in immature individuals. TAXONOMY The characters found to be of taxonomic importance among the species of Tantilla in the southeastern United States are: (1) ventrals, (2) subcaudals, (3) tail to total length ratio, (4) degree of develop- ment of the light parietal band, (5) width of the black nuchal collar - posterior to the parietal area, and (6) number of basal hooks of the hemipenis. Other characters recorded that either do not vary or do so without demonstrable geographic correlation are the following: (7) number of dorsal scale rows, (8) preoculars, (9) postoculars, (10) temporals, (11) supralabials, (12) infralabials, (18) number of supra- labials entering the orbit, (14) number of infralabials contacting the anterior chin shields, (15) condition of the anal plate, (16) contact be- tween the mental and anterior chin shields, (17) number of maxillary teeth, (18) length of retracted hemipenis, (19) degree of pigmentation of the postoculars and ultimate supralabial, (20) lateral extent of the 1966 VARIATION AMONG SOUTHEASTERN SNAKES 265 black nuchal collar, and (21) initial and terminal vertebral scales con- tacted by the black nuchal collar. Chatacters 1 to 21 were recorded for ali populations. THE PENINSULAR POPULATIONS. Populations of Tantilla from peninsular Florida are presently des- ignated as Tantilla coronata wagneri (Jan) The holotype of Homalocranion wagnefi was 6riginally located, at Munich, according to Jan and Sordelli (1866). Dr. Walter Hellmich indicated (in litt.) that it is not there and was perhaps destroyed dur- ing World War II. Dr. M. Torchio of the Museo Civico di Storia Naturale, Milan, which had many of Jan's types, has informed me that the Museum was destroyed during World War II and that if the type was there rather than at Munich, it has not survived. In lieu of examining the holotype, it is necessary to rely upon the descrip- tion, and the fortunate illustration of the presumed type of Jan and Sordelli. I have compared Blanchard's translation of Jan's descrip: tion with the original and find no discrepancies. Consequently refer- ences to the type descriptiotn refer to Blanchard's translation. The holotype had "Behind two or three gular scutes ... 188 yentrals, an entire anal, and 45 double caudals." Disregarding the anal for the moment, when 188 ventrals are plotted against 45 cau- dals in scatter diagrams as in figure 1, it is obvious that the specimen lies well outside the range of variation of peninsular populations (with the possible exception of the Dade County population), and well with- in the distribution of Tantilla coronata. As sex of the type specimen is not given, it is plotted on scatter diagrams of both sexes. On this basis I conclude that Jan had at hand a specimen of Tantilla coronata from a non-peninsular population. Though the type locality is given as "Florida", it should be pointed out that Florida once extended from the Atlantic Ocean to the Mississippi River along the Gulf Coast. Even today those Louisiana parishes east of the Mississippi are called the "Florida parishes." As Jan named the specimen in 1862, it was probably not collected much later than 1859. By that time very few American naturalists had collected in peninsular Florida, and I doubt seriously that Professor Wagner numbered_ among those who did. Tantilla coronata is common in the area between the Appalachicola and Mississippi rivers, and I suspect that Jan's specimen was from this area. Nothing in the type description settles the issue. Presence of a single anal is noteworthy, h6wever, for I found but two speci- mens among the 881 TantiIla I examined that had an undivided anal. A Tonlill. ..ron'll 67 Pen'.sul.r P.pul.liolls .. Males • •A . 0 0 8 . 7 O Seahors. Key ..: e A'B hvkids 60- ..... . C V D -I I j h.1.I,pl T.¢oronolo wignoil 0 4 :t su bc au da ls su bc ou da ls 0 0 0• A v I . 000. le .A *A A - e Abl A 0.0 % A A+E.r A 4q_ *A A A A 30 i Z i I I 110 120 vent,615 140 150 67- Females 60 . . .. 0 e. ® 0 .,JA * 00 0. 0 A /Vwb 60 .... 0 **A.0 AA A o tri. t:*WA A 40 AA* , .AA - 0 1= AA I . A A AA A A 30 11'0 110 11vent,als 140 1~0 Fig, 1. Scatter diagrams plotting ventrals against subcaudals. Enlarged symbols represent population means, except where other.wise indicated. 1966 VARIATION AMONG SOUTHEASTERN SNAKES 267 The illustrati6n in Jan and Sordelli of H. wagneri is that of a typical Tantilla coronata with a prominent parietal band. On the basis of the evidence cited, I place the name Homalocranion wagneri Jan, 1862 in synonymy of Tantilla col'onata Baird and Girard, 1858. While Blanchard was correct in stating that overlap in scale characters between peninsular Tantilla and T. coronata exists, this overlap is evident only when ventrals or subcaudals are considered separately. If viewed as a combination of characters, as in the seat- ter diagrams (figure 1), specimens from the peninsula, except for those from Dade and Monroe counties on the extreme southern periphery of the range, show no appreciable Overlap with T. coro- nata: 4.1 percent of males and 6.1 percent of females from peninsular populations lie within the distribution of T. coronata on the scatter diagrams. Conversely, only 1 percent of males and 2.6 percent of females of T. coronata lie within the distribution of peninsular Tan- tilla in figure 1, again with the exception of Dade and Monroe Coun- ties. I interprete this to mean that no intergradation exists between T. coronata and peninsular populations of Tantilla, which the distri- bution map (figure 2) substantiates. No specimens of Tantilla are available from a broad area between the Appalachicola and Suwan- nee rivers extending north to Chattahoochee and Irwin Counties, Georgia. Though the hiatus may be real, it may also reflect lack of collecting, for apparently suitable habitat occurs sporadically through- out this area. The scatter diagrams also reveal substantial scutellation differ- ences between Dade County Tantilla and other peninsular popula- tions, and affinity between Dade County Tantilla and T. coronata. A hiatus of approximately 50 miles exists between the northernmost locality for Dade County specimens at Miami and the southernmost site for Tantilla at Boynton Beach, Palm Beach County. There is no evidence of intergradation between these two populations. Hemipenial characters of all the main populations of southeastern Tantilla lend support to the conclusions drawn from characters of scalation. The hemipenis of Tantilla (figure 3) is single, bearing a simple sulcus spermaticus, and ornamented with undifferentiated spines. In Tantilla coronata the basal half is naked, the distal half spinose, with spine size decreasing toward the apex, which is covered with small spines. There are two basal hooks, a larger postero-lateral hook adjacent to the sulcus in the basal third, and a slightly smaller antero-medial hook in the mesal third of the organ. The hooks are well differentiated from the larger spines, being a good third longer and half wider than the basal spines in the apical half. Although 268 BULLETIN FLORIDA STATE MUSEUM Vol. 10 the spines show minor differences in extent and size, the hemipenes of penihsular populations from the Suwannee River south through Palm Beach County generally resemble that of T. coronata-with one important difference: only the large postero-lateral basal hook, adjacent to the sulcus, is present. The antero-medial hook has ap- parently been lost. Two basal hooks at approximately the same size are present in the Dade County population. A Al,bi- ..... A Foll line D Okele,okee Swamp Appolachicola R. Suwonnee R. .r. Lake George Seahorse Key -I :I ~ II 1 St. John'$ R. Hilliboro R. i.ke Wa/es Ridge Kissimmee R. peace R. ..0.. Fig. 2. Distribution of Tantila in relation to ayailability of sgub and sand- hills habitat in southeastern United States. TFiangles T coronata; circles main peninsular population; squares Dade County population. :*. ]141. .. - 6% 5 -Fl.j/...491. ·'1' .,y3*:1.E.1..'.: t 4,¥·'· 90 )1 ../.:.. 4- 94'9.. i. , . ~AY:.{i..)13..459{.f]>i,· ·: ·*f:}:".**·-& it·'~ .* r,i 95.·:.Z·Y Fig. 8. Hemipenes of southeastern Tantma: a, T. coronata; b, peninsular population A; c, peninsular population D; d, peninsular population C. 270 BULLETIN FLORIDA STATE MUSEUM Vol. 10 Other differences between these three populations exist, and will be discussed below, along with variation within the divisions. I be- lieve that the characters discussed so far, in conjunction with the real or apparent hiatuses in distribution, justify division of the southeast- ern Tantilla into three distinct species. Tantilla coronata extends from Florida north and west of the Suwannee River to Virginia, the Mississippi River, and extreme southern Indiana. A second species extends from the Suwannee River south in Florida to the Peace River on the west coast, and to Broward County along the east coast. The third is restricted to Dade and northern Monroe counties in extreme southern Florida. In recognition of its probable origin during fluctuations of Ple- istocene sea levels, I designate the populations of Tantilla within peninsular Florida south of the Suwannee River and north of Dade County and the outlet of the Peace River into the Gulf of Mexico as Tantilla relicta new species HOLOTYPE: UF 12421, an adult female collected 26 December 1960 on the south side of Babson Park, Polk County, Florida by Sam R. Telford, Jr. PARATYPES: Polk County: UF 12429-1&2, 12420, 12419; Palm Beach County: UF 12480-1&2; Alachua County: UF 12406, 7978- 1&2, 8586. DIAGNOSIS: A small Tantilla, characterized by the presence of a single basal hook on the hemipenis, postero-lateral in position and adjacent to the Sulcus spermaticus in the basal third of the organ. The species is composed of several populations varying somewhat among themselves in details of pattern and scutellation, divisible into three distinct subspecies. Ventrals range between 115 and 142; sub- caudals from 89 to 67. Males have fewer ventrals and more subcau- dals than females, and are slightly smaller. Although constant with- in subspecies, head pattern varies from completely black with no light parietal band, to a prominent parietal band separating the black head cap from the black nuchal collar, with extensive unpig- mented areas extending forward along outer edges of parietals. Ratio of tail to total length varies between 18.5 and 29.1 percent, males having longer tails on the average. Maxillary teeth 14 to 16, usually 15. Head pointed in outline with countersunk lower jaw, to nar- rowly rounded with noncountersunk lower jaw, varying somewhat among populations. DESCRIPTION OF HOLOTYPE: Adult female, f 186.5 mm .snout-vent 1966 VARIATION AMONG SOUTHEASTERN SNAKES 271 and 84.5 mm tail length. Ventrals 181; subeaudals 50; dorsal'scales smooth, in 15 rows throughout; on each side are 1 preocular, 2 post- oculars, 1 anterior and 1 posterior temporal, 7 supralabials, and 6 infralabials. Supralabials III and IV enter the orbit; infrajabials I to IV contact the anterior chin shields; mental plate in contact with anterior ehin shields; posterior chin shields are two-thirds lenkth,of anterior; rostal slightly scarred; anal divided. In preservative the head dorsum is black. A light parietal band, about one and one-half dorsal scales in width, extends from the rear fourth of parietals across the distal half of the secondary temporals; the light band is broken by a dark streak extending along the mesal edges of the parietals and the parietal suture to unite with the black Erst vertebral scale. The black nuchal collar is three and one-half vertebral scales in width and extends laterally on to scale row 2; supralabials I to IV and VI, the loreal region, and superior postocular are black; the distal fourth of inferior postocular, anterior fourth of primary temporal, and dor- sal edge of supralabial V lack pigment; the distal half of supralabial VII lacks pigment, the light area merging with the parietal band. The mental, infralabials I to V, and adjacent edges of the anterior chin shields are suffused with black; infralabial VI has dark fiecking. The dorsum is light tan, and the venter tannish white. In life, the dorsum was light tan (army brown, 6-A-10) throughout, becoming lighter laterally to pinkish tan (rosestone, 6-C-9). Tail venter pink- ish gray (cobwebs, 5-B-7), becoming white along body venter. Parie- tal band tan (French beige, 13-A-7); head dorsum dark brown (sepia. 8-A-10); nuchal collar black. VARIATION IN T~XONOMIC CHARACTERS. The correlation and distribution of variation within Tantilla relicta show the total sample to be composed of several distinct groups which should be discussed individually. The three most distinct groups of specimens are treated below as subspecies, although no btrong evi- dence is available to establish continuity of gene flow among 'them, For the moment they are designated populations A, B, and C. Population A extends from the Suwannee River south to the Hills- borough River along the west coast of Florida, and in the central por- tion of the state, west of the St. Johns River to northern Polk County, It is apparently found only in sandhills and mesic hammock habitats. Population B occurs in disjunct demes from the vicinity of Lake George south along the central ridge to southern Polk County (and presumably Highlands, as well), and in insular type situations along 3TABLE 1. VARIATION OF VENTRALS AND SUBCAUDALS AMONG SOUTHEASTERN Tantilla POPULATIONS. Ventrals Subcaudals N range mean SI N range mean 5 T. coronata males 109 128-140 182.5 10.8 97 84-58 46.7 95 females 86 188-147 140.0 10.2 77 85-50 43.1 9.0 T. oolitica inales 6 187-148 138.3 5.4 4 51-63 55.3 28.3 females 5 185-146 140.2 19.8 8 45-48 46.8 2.5 T. reticta B U LLE T IN FLO R ID A S TATE M U S E U M Vol. 10 Pop. A males 50 123-185 127.5 6.9 45 51-67 59.1 10.7 females 28 129«142 184.2 9.6 20 46-60 58.8 18.1 Pop. B males 8 117-181 128.6 22.9 8 44-59 50.8 26.1 females 14 120-184 128.0 19.2 10 40-55 48.1 22.8 Pop. C males 7 115-118 116.3 2.8 5 45-51 47.8 7.8 females 10 119-129 128.2 8.0 5 45-51 48.2 7.3 Pop. D males 8 121-184 126.8 18:7 7 48-52 50.6 2.0 females 11 128-189 188.2 40.4 8 45-54 50.8 10.1 Seahorse males 6 119-125 122.5 4.6 6 45-55 50.2 13:8 females 5 128-184 180.6 5.8 4 50-51 50.5 0.8 1966 VARIATION AMONG SOUTHEASTERN SNAKES 278 the west coast from Seahorse Key in Levy County to Charlotte Coun- ty north of the mouth Of the Peace River. In areas of sympatry with population A along the northern ridge, B is restricted to scrub habi- tat, but in aliopatric areas of the southern ridge, 'it also occurs in sandhills. Population C is found in coastal dunes and scrub from central Brevard County to southern Palm Beach County. The southern- most record is some 50 miles north of the distinctive Dade County population. Population D is comprised of 19 specimens that cannot be as- signed satisfactorily to either A or B, although they were collected from scattered localities within and on the periphery of the ranges of both A and B. These specimens show an overlap of characters from both A and B and may represent intergrades. They are the prin- cipal reason fof considering A and B conspecific. The Seahorse Key population of T. relicta resembles population B in characters of scutel- lation and body proportions, but differs' somewhat in patterns; it is therefore treated separately in the tables and figures. Variation in ventrals and subcaudals is presented in Table 1. The Dade County population evidently has the highest number of ventrals among these popullitions, while the T. relicta population closest to it geographically, C, has the lowest. T. coronata has the fewest subcaudals, and the T. relicta population geographically closest, A, has the highest number. It is tempting to speculate that these are examples of character displacement. Variation in proportionate tail length is considerably less than in scale characters. Table 2 presents the ratio of tail length to total length among the various samples, while figure 4 depicts regression lines for this character, calculated by the method of Snedecor (1946: 108). In correlation with subcaudal number, population A again has the proportionately longest tail, while T. coronata has the shortest. As the sexes showed no significant differences in color pattern, the data were lumped for analysis. Sorting specimens according to the distribution of pigment in the parietal and temporal regions permitted dividing them into ten different categories (figure 5) rang- ing from specimens with a completely black head cap which fused with the nuchal collar (category 1) to specimens in which only the anterior, central portion of the parietal was pigmented (category 10). This procedure provided a simple method of quantifying the pat- tem data (table 8). 274 B U LLE TIN FLO R ID A S TATE M U SEU M Vol. 10 TABLE 2. TAIL LENGTH IN SOUTHEASTERN Tantilla EXPRESSED AS PERCENT OF TOTAL LENGTH. Males Females N range mean S' N range mean f T. coronata 94 17.2-23.0 20.6 1.4 77 15.9-21.2 18.2 5.4 T. oolitica 4 20.8-22.4 21.9 0.6 8 19.1-19.6 19.8 0.1 T. felicia Pop. A 45 21.7-28.7 25.0 2.1 20 19.8-29.1 22.8 4.6 Pop. B 7 20.9-28.5 22.2 1.2 10 18.5-28.1 20.9 2.1 Pop. C 5 21.7-24.0 22.7 1.8 5 18.7-28.2 21.1 2.8 Pop. D 7 20.7-28.8 21.9 1.7 8 19.0-22.7 21.2 1.9 Seahorse 6 21.5-25.2 28.0 1.8 4 26.6-21.6 21.2 1.0 1966 VARIATION AMONG SOUTHEASTERN SNAKES 275 70- b d lo lle n g ,h ( m m ) lo il le ng ,h (m m ) Males 2- 1 1 8 192 70- b d Females 2- 1 1 snoubvent length (mm) /92 Fig. 4. Regression lines depicting the relationship of tail length to snout- vent length: a, peninsular population A; b, peninsular population B; c, T. core- nata; d, peninsular population D. 276 BULLETIN FLORIDA STATE MUSEUM Vol. 10 1. abcd 1 1 efgh i j Fig. 5. Variation ift parietal pattern among southeastern Tantilia popula- tions. Categories a-d are non-banded, e-j banded. 1966 VARIATION AMONG SOUTHEASTERN SNAKES 277 TABLE 8. PARIETAL PATTERN* IN SOUTHEASTERN Tantilla POPULATIONS, EXPRESSED AS PERCENT OF SAMPLE Pattern Types Nabedefghijmean T. coronata 117 3 16 87 28 9 9 8 6.6 T. oolitica 10 80 10 10 1.7 T. relicta POP. A 77 20 27 39 14 2.5 Pop. B 28 14 56 22 4 4 5.4 Pop. C 18 83 6 22 22 17 7.5 Pop. D 19 16 42 32 5 5 3.4 Seahorse 11 9 18 9 86 18 9 3.7 * See Fig. 5. Somewhat arbitrarily, the dividing line between banded and non- banded patterns was set between categories 4 and 5, with pattern types 5 and above considered banded. According to this classifi- cation population A is clearly 100 percent nonbanded, while 87 percent of B, 100 percent of C, and. 96.6. percent of T. coronata are banded. The small sample from Dade County (9) are all non-banded; a single specimen from Key Largo has a broken band. Of the 11 snakes in the small Seahorse Key sample, 8 are nonbanded, as are 2 of 4 T. relicta from Charlotte County. The series of possibly inte- gradient T. relicta, D, whose body proportions and scutellation char- acters are perhaps closer to population B, are 89.5 percent non- banded. Another Dattern character of taxonomic use is the width of the black nuchal collar at midline. The data are readily quantified by expressing the width in terms of number of vertebral scales pig- mehted (table 4). The banded populations, T. coronata, B, and C, have nuchal col- lars predominantly four scales or less in width, the percentages of the respective samples being 99.5, 91.5, and 67.0. Of the nonbanded populations, A and Dade County, 66 percent of A have collars five scales or more in width, while 4 of the 10 Dade County specimens have collars of this width. Eight of the 11 Seahorse Key Tantilla relicta and 52 percent of the possibly intergradient population, D, have collars Bve scales or more in width. Again, with respect to both pattern components discussed, the greatest contrast is between T. coronata and the T. relicta population closest to it geographically, population A. 278 BULLETIN FLORIDA STATE MUSEUM Vol. 10 TABLE 4. WOTH OF NUCHAL COLLAR IN SOUTHEASTERN Tantilia POPULATIONS, EXPRESSED AS PERCENT OF EACH SAMPLE Width in Vertebral Scales N12845678 9 mean T. caronata 183 18 63 28.5 0 .5 8 . 0 T. oolitica 10 10 20 80 20 20 4.2 T. relicta Pop. A 76 16 18 46 9 8 8 4.8 Pop.B. 23 4 17 26 44 9 8.4 Pop. C 18 11 89 17 83 8.7 Pop. D 19 11 5 16 16 47 5 4.1 Seahorse 11 18 10 27 27 18 5.8 In summary of Variation in the five characters discussed above, it is evident that the three main populations of T. relicta are better distingufshed among themselves, as well as from T. coronata and the Dade County population, by parietal pattern and width of the nuchal collar than by characters of scutellation and body proportion. These latter, however, are useful in some instances, i.e., the longer tail and greater number of subcaudals of population A in comparison to other populations, and the lower number of ventrals in an admittedly in- adequate sample of population C. I believe that sufficient grounds exist to designate populations A, B, and C as subspecies of Tantilla relicta, representing geograph- ically and ecologically delimited populations that possess within themselves common and distinctive morphological features. In accordance with this concept of the subspecies, I designate population B Tantilla relicta relicta new subspecies HOLOTYPE: UF 12421, an adult female collected 26 December 1960 on the south side of Babson Park, P6lk County, Florida by Sam R. Telford, Jr. PARATYPES: UF 12429-1&2, 12420, 12419, 12418, 12424; USNM 28429, 28480; CNHM 29592, 29595. DIAGNOSIS: A population of Tantilla relicta composed of many more or less isolated demes, restricted in northeentral Florida to scrub habitat, but occurring also in sandhills of southcentral Florida, where syntopic to other subspecies of T. relicta. Ventrals range from 117 to 184; subcaudals between 40 and 59. Tail length varies 1966 VARIATION AMONG SOUTHEASTERN SNAKES 279 from 18.5 to 23.5 percent of total length. Parietal pattern is predom- inantly banded (87 percent). The nuehal collar is usually (48.5 per- ,cent) 4 scales wide at midline, but ranges from 1 to 5 (mean, 8.4 scales). The head .is pointed in outline, with countersunk lower jaw. GEOGRAPHICAL RANGE: Peninsular Florida from the vicinity of Lake George, Marion County, south along the central ridge to south- ern Polk County. Presumably this subspecies occurs in Highlands County as well. Disjunct populations occur in coastal scrubs of Charlotte, Sarasota, and Pinellas counties, and on Seahorse Key, Levy County. DESCRI]mON OF HOLOTYPE: The holotype is described above as holotype of Tantilla relicta sp. nov. In recognition of Wilfred T. Neill's contributions to the bi6graphy of Florida, I take pleasure in designating the nonbanded population of Tantilla felicta in northcentral Florida (A) as Tantilla relicta neilli new subspecies HoLoTypE. UF 12406, an adult male collected 12 July 1961 on the University of Florida campus, Gainesville, Alachua County, Flor- ida, by John Funk. PARATYPES: UF 7978-1&2,3586, 12404,8980, 1047, 9586; CNHM 8558; UMMZ 44967; MCZ 48180. DIAGNOSIS: A population of Tantilla felicia relatively continuous in distribution in habitats of sandhills and mesic hammock of north- central Florida. Ventrals range from 128 to 142; subcaudals between 46 and 67. Tail length varies from 19.8 to 29.1 percent of total length, and is proportionately longer in this subspecies than in any other population of Tantilla in southeastern United States. The parietal pattern is nonbanded in all specimens examined, with only 15 per- cent having unpigmented areas on the parietals larger than a dorsal scale in size. The black nuchal collar, which fuses with the black parietal area, is usually (46 percent) 5 scales wide at midline, but varies from 3 to. 8 (mean, 4.8 scales). The head is less sharply pointed in outline than that of T. relicta relicta, and the lower jaw is less noticeably countersunk. GE6GRAPHICAL RANGE: Peninsular Florida from the Suwannee River south to the Hillsborough River and northern Polk County, and east to the St. Johns River. DESC*UPTI ON OF HOLOTYPE: Adult male, 168.0 mm snout-vent and 52.0 mm tail length. Ventrals 181; subcaudals 60; dorsal scales smooth, in 15 rows throughout; on each side are one preocular, two 280 BULLETIN FLORIDA STATE MUSEUM Vol. 10 postoculars, one anterior and one posterior temporal, seven suprala- bials, and six infralabials. Supralabials III and IV enter the orbit; infralabials I to IV contact anterior chin shields; mental plate sep- arated from anterior chin shields by infralabials I; posterior chin shields are two-thirds length of the anterior; anal plate divided. In preservative the head dorsum is black with no light parietal band; an area the size of one dorsal scale on the posterior tip of the parie- tals lacks pigment. The black nuchal collar, fused with the black of the parietals, extends onto the anterior tip of vertebral scale IV, and laterally onto scale row 2 of each side. The mental, infralabials I to V, and lateral half of the anterior chin shields are heavily suf- fused with black; infralabial VI and the gular scales have some black flecking. The entire loreal and postocular regions and supralabials I to VI are black; supralabial VII is black except for a small light area on the rear ventral third. The dorsum is dark tan, venter white. In life the dorsum was dark tan (argus brown, 7-A-12) becoming lighter (Natal brown, 7-A-10) on the sides. The tail venter was yellowish cream (amber white, 11-C-1), and the body venter pinkish gray (greystone, I2-A-2). The head dorsum and gulars were black, with the gulars a slightly lighter shade. The parietal spots were dark tan (Vandyke brown, 7-A-11). ' The third population of T. relicta, C, inhabits coastal dunes and scrub along a narrow strip of Pamlico tetrace in southeastern Florida. In reference to the geological origin of its habitat, I designate this population Tantilla relicta pamlica new subspecies HoLorYPE: UF 12430-2,- an adult female collected 4 February 1960 approximately 1 mile south of Boynton Beach, Palm Beach County, Florida, by Robert H. Mount. PARATYPES: UF 12480-1, 12423; MCZ 19146, 16272, 16278, I2802; UMMZ 55779-1&2, 84198. DIAGNOSIS: A population of Tantilla relicta occurring in.relatively isolated coastal dunes and scrub of southeastern Florida. Ventrals range from 115 to 129 and are fewer in this subspecies than in any other southeastern Tantitta population. Subcaudals range between 45 and 51, and tail length varies from 20.7 to 24.0 percent of total length The parietal pattern fs banded in all specimens examined; over 60 percent show a trend t6ward total absence of pigment on the parie- tal, supraocular, and temporal regions, a reduction in pigmentation found in less than 10 percent of T. r. relicta. The nuchal collar is usually (40 percent) 8 scales in width, and ranges from 2 to 5 (mean, 1966 VARIATION AMONG SOUTHEASTERN SNAKES 281 3.7,scales). The head is prominently pointed in outline, with counter- sunk lower jaw. In life the dorsum is reddish brown, a color seldom observed in T. r. relicta. GEOGRAPHICAL RANGE: The east coast of Florida from the vicin- ity of Cape Canaveral south to southern Palm Beach County, and limited to the west by the Kissimmee River valley. DESCRIPTION OF HOLOTYPE: Adult female, 166.0 mm snout-vent length, with an incomplete tail 81.0 mm long. Ventrals 120; subcau- dals, incomplete) 30; scales smooth, in 15 rows throughout; on each side are one preocular, two postoculars, one anterior and one posterior temporal, seven supralabials, and six infralabials. Supralabials III and IV enter the orbit; infralabials I to IV contact the anterior chin shields; mental plate is in contact with the anterior chin shields, which are one-third longer than the posterior chin shields; anal plate dividedf In preservative the frontal, supraoculars, and anterior two- thirds 62 the parietals are black; the unpigmented posterior third forms a parietal band, which is narrowed slightly at midline by the black vertebral scale I; the nuchal collar is black and extends onto the anterior edge of vertebral scale V, and laterally on both sides onto scale row 2. The medial anterior corners of the secondary tem- poral are black, the remainder unpigmented. The loreal region, su- pralabials I to IV, and superior postocular are black; pigment is lack- ing from the rostral, three-fourths of the internasals, and nasal half of the prefrontals, anterior edge of the frontal, antero-ventral two- thirds of the primary temporals, distal edge of the inferior postocu- lar, supralabials V to VII, and infralabials V and VI. Infralabials I to IV have a light suffusion or flecking of black. The dorsum is light pinkish tan, and the venter pinkish white. In life the dorsum was light reddish brown (cedarwood, 6-G-10), becoming lighter some- what (plantation, 6-D-11) in the anal region. Sides were pinkish tan (roseglow, 5-D-9). The tail venter was pinkish gray (cobweb, 5-B-7), becoming lighter (iris mauve 8-8-7) on the body venter. The parietal band was a prominent tan (suntan, 18-8-7), while the head dorsum was dark brown (bracken, 8-C-12), and the nuchal collar Black. I designate the distinctive species of Tantilla restricted to Dade and Monroe counties, Florida, after its association with the dominant geological formation of the area, the Miami oolite, as Tantilla oolitica new species HoLOTYPE: UF 17826, an adult male collected in April 1955 in Miami, Dade County, Florida in a vacant lot on southwest 27th 282 BULLETIN FLORIDA STATE MUSEUM Vol. 10 Avenue near 24th Street. Collect6r unknown, received through the courtesy of Dennis R. Paulson. PARATYPES: Dade County, UF 17825, MCZ 87899; CM 20077; UM 55-137, 55-686; Monroe County: UMMZ 108174. DIAGNOSIS: A species of Tantilla resembling Tantilla coronata in details of hemipenial structure and scutellation. The hemipenis has both basal hooks, the postero-lateral hook adjacent to the sulcus spermaticus in the basal third, and the antero-medial hook, approxi- mately equal in size to the postero-lateral, in the mesal third of the hemipenis. Ventrals range from 185 to 146; subcaudals between 45 and 63. Males appear to have more subcaudals than females; too few specimens are available to demonstrate a trend in ventrals. Females exceed males in size. Two types of head pattern are evi- dent; all specimens from Dade County exhibit a completely black head and neck with no light parietal band; a single specimen from Key Largo, Monroe County, has a prominent but broken band sep- arating the black head cap from the black nuchal collar. Males have slightly longer tails than females, with tail length ranging (both sexes) from 19.1 to 22.4 percent. Maxillary teeth 14, 15 or 16. Head broad- ly rounded, with noncountersunk lower jaw. DESCRIPnON OF HOLOTYPE: Adult male, 166.5 mm snotit-vent and 48.0 mm tail length. Ventrals 137; subcaudals 68; dorsal scales smooth, in 15 rows throughout; 1 preocular, 2 postoculars, 1 anterior and 1 posterior temporal, on each side. There are 7 supralabials on the left and 6 on the right, with III and IV entering the orbit on both sides. Infralabials are 5 on the left and 6 on the right; I to III contact the anterior chin shields on the left, and I to IV on the right side. Mental plate contacts the anterior chin.shields; posterior chin shields two-thirds length of the anterior. In preservative the head dorsum is black; no light parietal collar, black of the head extending along the neck to the middle of vertebral scale IV and to scale row 2 lat- erally. Mental and infralabials heavily suffused with black, as is the rear half of infralabial VI at angle of the jaws. Supralabials I to IV and postoculars completely black, as is loreal region. Supra- labials V and adjacent portions of VI and primary temporal lack pigment; supralabial VII largely black, lacking pigment in most of rear half. Dorsum is light tan, and venter white. No color notes from life ayailable. THE NON-PENINSULAR POPULATIONS. Schwartz (1958) based his diagnosis of Tantilla coronata mitrifei' upon three main points: 1966 VARIATION AMONG SOUTHEASTERN SNAKES 288 (1) .posterior ventrad extension of the black cap tends not to reach the upper labial border..." (2) «black collar does not extend farther posteribrly than the fourth scale row ib the midline..." (8) "with more ventral and less subcaudal scutes on the aver- age... I have examined the holotype from Caesar's Head, Greenville Countyi South Carolina (CHM 58-92.2), and six paratypes from Roundtop Mountain, Rutherford County, North Carolina (AMNH 66207-66211). Other paratypes were not made available to me. The holotype certainly agrees with the main points of the subspe- cies diagnosis quoted above, but of the six paratypes I examined, only one agrees with the holotype in character (1) in the other five, the ventrad extension extends well past the upper labial border. This Variation appears commonly in other parts of the range; speci- mens from Indiana, Edmonson County, Kentucky, and Jackson Coun- ty, Florida, lack a ventrad extension to the black cap. Character (2) also lacks exclusiveness: one of the six paratypes has the black nuehaI collar extending on to vertebral scale V, although in the other para- types and holotype the collar terminates on scale IV. In a series of 87 T. coronata from coastal plain Mississippi, 62 percent (28) have the nuchal collar ending on scale IV or anteriorly; in 38 percent (14), it extends onto scale V. Thus, it appears that two of the main char- acters used by Schwartz are not sufficiently exclusive to be of diag- nostic value. This leaves only characters of (8) to be evaluated. Ventrals and subcaudals of 5 female T. coronata mitrifer types were compared with a series of 14 female coastal plain T . coronata coronata from Charleston and Berkeley counties, South Carolina, specimens which were examined by Schwartz. Admittedly five speci- mens is an inadequate sample, but standard tests (the Student T) were performed anyway, with the following results: there is a difference in means of ventrals and subcaudals between coastal plain females (188.8,42.5) and T. coronata mitrifer females (142.8, 89.5). The difference in ventrals is significant at the .05 level; that in sub- caudals is not significant at the same level. Schwartz states that T. coronata mitrifer has a darker dorsum than does T. coronata coronata. I observed that in the few living T. coronata I have seen those specimens from the Upper Piedmont were considerably darker than those from the Coastal Plain. In many preserved specimens this distinction is still remarkably evident, and 284 BULLETIN FLORIDA STATE MUSEUM Vol. 10 specimens from the Blue Ridge and Upper Piedmont constitute a distinct group from those of the Coastal Plain. It seemed reason- able that two distinct populations of Tantilla coromita might exist, those of the Coastal Plain and those Schwartz called T. coronata mitrifer, but including as well demes occupying the Upper Piedmont. The region of intergradation might logically follow the Fall Line. With this hypothesis in mind, I segregated Tantilla from N6rth and South Carolina, eastern Tennessee, Georgia, Alabama, Mississippi, and Louisiana into three groups: Blue Ridge and Upper Piedmont, Lower Piedmont and Fall Line, and Coastal Plain (table 5). TABLE 5. SCALE CHARACTERISTICS OF TantiIla coronata POPULATIONS Upper Piedmont Coastal Plain Fall Line and Montane mean ventrals males 131.6 184.2 132.0 females 188.1 141.9 142.2 mean subcaudals nnaIes 46.9 46.5 46.8 females 48.5 44.4 42.6 mean proportionate tail length (%) nnales 20.7 20.5 20.1 females 18.6 18.5 18.8 mean collar width in number of scales males 3.1 2.9 2.6 females 3.1 3.1 2.8 distal vertebral scale included in collar (% of sample) males 8 1.8 11.1 11.0 4 50.0 55.6 89.0 5 48.2 38.8 females 8 11.8 4 65.8 57.1 70.6 5 81.6 42.9 17.6 6 2.6 If one postulates that T. coronata mitrifer, in a wider sense geo- graphically than Schwartz defined it, is characterized by more ven- trals, fewer subcaudals, a narrower nuchal collar that tends to lie 1966 VARIATION AMONG SOUTHEASTERN SNAKES 285 more anteriorly on the neck, and a shorter tail, in combination with a darker dorsum, in comparison with T. coronata from the Coastal Plain, then the data cited tend to support each of these distinctions. And in almost every character, specimens from the proposed region of intergradation are intermediate between mitrifer (sensu latu) and coronata (sensu strictu). Standard statistical comparison (the Student T test) was made for each of these characters. Differences in mean number of ventrals of both sexes, position of the band (data lumped, as males and females show no significant differences) and collar width in males are significant at the .05 level. Differences in mean number of subcaudals and proportionate tail length in both sexes, and collar width in females are not significant at .05. On the basis of this analysis, I think a weakly defined population of Tantilla coronata exists in montane regions and the Upper Pied- mont of North and South Carolina, Georgia, eastern Tennessee, and Alabama, to which the subspecific designation Tantilla coronata mitrifer Schwartz might be applied. I consider the variation in scale characters, however, to be insufHcient to warrant such distinction. Although a striking contrast in coloration of the dorsum apparently exists, as illustrated by color descriptions below, living material avail- able to me has been too meager to substantiate the value of this char- acter. At this time, I think it is best to relegate the subspecies T. coronata mitrifer to the synonymy of Tantilla coronata Baird and Girard, and recognize no subspecies of this wide-ranging species. It is appropriate at this point to deRne my concept of the species Tantilla coronata, as developed from examination of 215 specimens from all areas of its range. Tantilla coronata Baird and Girard, 1853 HOLOTYPE: USNM 1876, collected in Kemper County, Mississip- pi, by D. C. Lloyd. DIAGNOSIS: A species of Tantilla characterized by the presence of two basal hooks on the hemipenis, one posterolateral in position and adjacent to the sulcus spermaticus in the basal third of the organ, and the Other, slightly smaller and anteromedial in position, in the mesal third of the hemipenis. Ventrals range from 128 to 147; sub- caudals between 84 and 53. Males have fewer ventrals and more subcaudals than females, and longer tails. Tail to total length ratio varies from 15.9 to 28.0 percent. A prominent light parietal band is present in most (96.6 percent) specimens, separating the black cap of the head from the black nuchal collar. The nuchal collar is usu- ally (76 percent) 3 or fewer scales in width at midline, ranging from 286 BULLETIN FLORIDA STATE MUSEUM Vol. 10 2 to 6 (mean, 3.0). On each siae are usually one preocular, two postoculars, one anterior and one posterior temporal, seven supra- labials, and six infralabials. Supralabials III and IV enter the orbit; infralabials I to IV contact the anterior chin shields, which may or may not be separated from the mental plate by infralabials I. Pos- terior chin shields two-thirds length of anterior. Anal plate divided. Maxillary teeth (one side) 14 to 18, usually (55 percent) 16. Dorsum ranges from tan in coastal plain populations to dark brown in mon- tane localities; venter white. GEOGRAPHICAL RANGE: T. coronata is found from Florida west of the Suwannee River westward to the Mississippi River and north to southern Indiana and Virginia. A specimen of T. coronata from Coffee County, Alabama (UF 12887) was obtained within a day of its preservation, and the follow- ing color notes taken. The dorsum was uniform light brown (army brown, 6-A-10), becoming tan (rose blush, 5-C-9) laterally. The ven- ter ranged from cream (9-D-2) beneath the tail to pinkish white (tilleul-buff, 10-A-2) anteriorly. The nuchal collar was black, pre- ceded by a gray-brown (manon, 6-A-9) parietal band. The black dorsal surface of the head faded to gray (82-A-1) on internasals and prefrontals. This specimen is representative of Coastal Plain popu- lations. Another specimen (UF 12409) collected in Cobb County, Georgia presents a vivid contrast to the Coastal Plain population. In life the dorsum was dark brown (sepia, 8-A-10) distally, becoming slightly lighter (moose, 8-C-10) at midbody, and light brown (bark, 8-C-11) on the anterior third of the body. The sides of* the body were dark tan (bracken, 8-C-12). The pinkish gray (rosedust, 6-8-2) tail venter became gradually lighter to pinkish white (folkstone, 18-A-8) on the anterior third. The parietal band was cream (vanilla, 10-C-8), and the nuchal collar and dorsal surface of the head were black. Identi- cal coloration was present in a series of three T. coronata (UF 12402) from 2 miles south of Millry, Washington County, Alabama, well with- in the Coastal Plain. In scalation and pattern, these specimens clearly be16ng to Coastal Plain T. coronata. As the Black Warrior River system flows from the Piedmont to the Gulf of Mexico through this region, it may well be that some gene flow from Upper Piedmont pop- ulations, which extend south as far as Tuscaloosa, follows this drain- age system well into the Coastal Plain. Tantilla coronata extends up the eastern Missis5ippi Valley through western Tennessee and Kentucky to southern Indiana. The Indiana series does not differ appreciably from Coastal Plain Mississippi speci- TABLE 6. SUMMARY OF VAIUATION IN CHARACTERS LACKING TAXONOMIC SIGNIFICANCE 1966 VA R IATIO N A M O N G SO U TH EASTER N SNAKES 287 % of each population varying from normal - · Normal Variations Character condition (No. cases) A B C D E F G preoculars 1-1 2-1 (1) 1-2 (1) 1.0 4.4 5.3 0-0 (2) postoculars 2-2 2-0 (1) 2-1 (2) 1.0 4.4 2.6 16.6 9.1 5.3 1-2 (4) 1-1 (8) supralabials 7-7 7-6 (7) 7-4 (1) 1.5 8.2 8.8 9.1 27..2 6-6 (4) supralabials III,IV-III,IV III,IV-II,III (1) in orbit II,III-II,III (3) 1.0 9.1 infralabials 6-6 7-7 (1) 7-6 (2) 6-7 (5) 6-5 (7) 6.2 8.7 '2.6 9.1 86.4 5-6 (2) 6-4 (1) 5-5 (8) TABLE 6. SUMMARY OF VARIATION IN CHARACTERS LACKING TAXONOMIC SIGNIFICANCE (continued) % of each population varying from normal 288 B U LLE T IN F LO R ID A S TAT E M U S E U M Vol. 10 Norrnal Variations Character condition (No. cases) A B C D E F G infralabials I-IV,I-IV I-IV,I-III (9) contacting I-III,I-IV (5) anterior I-V,I-IV (1) chin shields I-IV,I-V (4) 6.7 8.7 2.6 5.6 18.2 5.8 45.5 I-IV,I-II (1) I-III,I-III (4) I,II,IV,V;I-IV (1) I-III,V;I-V (1) temporals 1&1,1&1 1&1&1,1&1&1 (7) 1&1&1,1&1 (7) 1&1,1&1&1 (7) 6.2 4.4 9.0 11.1 .9] 1&1,0 (1) 1&1,1&2 (1) dorsal scale 15-15-15-15 17-15-15-15 (1) rows 16-15-15-15 (2) - 14-15-15-15 (2) 15-14-15„15 (1) 8.1 2.6 5.0 9.1 14-14-15-15 (1) 15-16-15-15 (1) 15-13-15-15 (1) 15-15-15-18 (1) anal plate divided single (2) 0.5 1.3 A. T.,coronata D . T. relicta pamlica G. T. oolitica B. T. relicta relicta E. Seahorse T. relicta C. T. reticta neilli F. relicta x neilli intergrades 1966 VARIATION AMONG SOUTHEASTERN SNAKES 289 mens. A single specimen was available from southern Virginia, and resembles those from coastal North Carolina. I have seen only two T. coronata from Florida, both from the vicinity of Marianna, Jackson C6unty, which. lies west of the Appalachicola River, and these re- semble the Alabama and Mississippi specimens. Although no speci- mens are in existence from the region between the Appalachicola and Suwannee rivers, south of Chattahoochee County, Georgia, T. coronata may be there. It does not cross the Suwannee River into peninsular Florida. VARIATION IN CHARACTERS OF NO TAXONOMIC SIGNIFICANCE. Most of the variation found in characters that lack taxonomic significance is presented in Table 6. Temporals and infralabials were found to be the most variable scales, and less than 10 percent of these varied from the usual condition in each population sampled, except in the T. relicta sample from Seahofse Key and the T. oolitica sample. Both samples numbered only 11 specimens each, and while small sample size probably influenced disproportionately the percent- age of specimens varying from the norm, the facts that both popula- tions are insular in nature and probably small in numbers may also be of moment. In the 11 T. oolitica specimens, 4 of the 8 showing scale aberrations were aberrant in multiple characters. Most cases of mul- tiple aberrations observed in all samples were correlations between variations in labial number and the number of supralabials entering the orbit or infralabials contacting the anterior chin shields. Table 7 presents the percentages of each sample that vary from the normal scutellation listed in Table 6. TABLE 7. PERCENTAGES OF ABERRANT SPECIMENS IN SOUTHEASTERN Tar:tilla POPULATIONS % aberrant % with in at least multiple N one character aberrations T. coronata 195 15 .4 48 . 4 relicta x neilli intergfades 19 15.8 1/8 T. relicta neilli 78 19 .8 26.6 T. relicta pan«ica 18 27 .8 1/5 T. relicta relicta 28 80 . 2 2/7 Seahorse T. relicta 11 86. 4 1/4 T. oolitica 11 73 . 0 4/8, all peninsular populations 165 25.4 81.0 290 BULLETIN FLORIDA STATE MUSEUM Vol 10 The peninsular populations clearly show a trend toward more aberrant individuals than does T. coronata. I attribute this to the distribution of peninsular Tantilla in discontinuous demes, with gene flow severely restricted by habitat barriers. The slightly lower per- centage of aberrant individuals of T. relicta neilli may reflect greater continuity of gene flow within the sandhills and mesic hammock habi- tats, in contrast to the insular natute of the occurrence of scrub. No significant differences were found between the sexes. Through an oversight on my part, one character, not listed above, was not recorded for all specimens examined. This is the contact of the mental plate with the anterior chin shields, and its corollary condition, separation from them by infralabials I. In view of the importance attributed to this character in some descriptions of Tan- tilla species, 58 specimens available to me at the time the oversight was discovered were checked to determine whether or not the char- acter might be taxonomically useful. Table 8 tabulates the results. Clearly, either condition may be encountered within the same population, and from this small sample, equal frequency is indicated. No correlations with sex were found. Descriptions of Tantilla species that stress this character should be re-evaluated. TABLE 8, CONTACT OF MENTALAND ANTERIOR CHIN SHIELDS IN SOUTHEASTERN 'Tanti~a POPULATIONS N Contact No contact T. coronata 22 15 7 T. oolitica 8 1 2 T. relicta relicta 6 3 8 T. reticta neilli 10 5 5 T. relicta pamlica 2 2 relicta x nemi intergrades 2 2 Seahorse T. relicta 8 2 6 Total 58 80 28 Maxillary dentition of several Tantilla species was described by Smith (1940: 61). In the group to which T. coronata belongs the two abruptly enlarged, grooved rear teeth are not in line with the anterior smaller teeth and are separated from them by a diastema. The 88 maxillae, I have examined in this study, 49 from T. coronata, 8 from T. oolitica, and 81 from T. relicta, all have a similar morph- ology. The diastema varies from one to two tooth spaces in width, 1966 VARIATION AMONG SOUTHEASTERN SNAKES 291 and the enlarged teeth are approximately three times the size of the preceding teeth. The smaller teeth decrease slightly in size as they approach the diastema, and are strongly recurved. Table 9 presents data on the total maxillary counts (enlarged plus smaller teeth) recorded. TABLE 9. NUMBER OF MAXILLARY TEETH (LEFT SIDE) IN ~OUTHEASTERN Tantilla ~OPULATIONS Number of Teeth 14 15 16 17 18 Mean T. coronata 6 16 26 1 15 . 8 T. ooMica 1 1 1 15.0 T. relicta relicta 2 6 5 15.2 T. reticta nei~ 3 6 4 15,7 T. relicta pamlica 2 3 14.2 T. relicta, all subspecies 7 15 9 15.4 Although T. coronata tends to have one more maxillary tooth (55 percent have 16 or more) than does T. relicta (71 percent have 15 or less), there is no significant difference in mean tooth count. Two few data were recorded concerning the position of the retracted hemipenis to evaluate this character properly, but the little information available suggests that it may be of little taxonomic use in this species group. The retracted hemipenis was found to extend to subcaudals 9 to 11 in T. coronata, 10 to 12 in T. relicta, and 8 in the single T. oolitica in which it was measured. The postero-lateral basal hook is located at subcaudals 8 or 4 in T. coronata, 4 or 5 in T. relicta, and 8 in T. oolitica. The antero-medial basal hook, ab- sent in T. relicta, is located at subeaudal 5 in both T. coronata and T. oolitica. Terminal spination begins at the level of subcaudals 5 to 7 in T. coronata, 7 or 8 in T. relicta, and 5 in T. oolitica. GROSS ECOLOGY Tantilla r. relicta is restricted, in areas of syntopy with T. r. neilli, to scrub habitat On St. Lucie fine sand. In areas of allopatry, T. r. relicta may inhabit sandhills, as at the type locality, Babson Park, or coastal dunes and scrub. T. r. neilli is found only in sandhills and dryer mesic hammock areas. T. r. relicta appears completely fossorial, while T. r. neilli, although semifossorial, may be found occasionally beneath logs, debris or rocks. Both subspecies inhabit the mounds 292 BULLETIN FLORIDA STATE MUSEUM Vol. 10 of loose sand thrust up by Geomys pinetis, the pocket- gopher. The distribution of both subspecids is confined to areas of dry soil in Flor- ida, as is that of T. r. pamlica in the coastal dunes and scrub (figure 2). Tantilla coronata, however, seems to have broader habitat tolerances, being found in sandhills, hardwood forests, and pine woods. I have been informed (J. Dobie, personal communication) that it is fre- quently collected from rotten stumps in seasonally flooded pine flat- woods of Mississippi and Louisiana. Collecting notes on 22 museum specimens include the fdllowing situations: rotten stump, in rotten pine log, in pine straw, under log in sandhills (8), under rocks (9), and in spider web! Habitat notes on these specimens include pine woods, pine-hardwood upland, edge of hardwoods, edge of river, and second growth of thin, oak-forested hillside. Neill (1951: 49) men- tions collecting T. coronata by "digging into piles of decaying bark sarap, by uprooting rotten stumps, or by overturning large rocks and boulders on the wooded hillsides," and finding the eggs in debris from a rotting longleaf pine in this .locality. Neill and Boyles (1957) discuss the eggs of a female collected "in leaf litter on a rocky ex- posure along the Black Warrior River" in Tuscaloosa County, Ala- Aug. 0 6 A A A •. 6• A - 6 0 0 A A*'Al A.A - %.5 AA* .. .006 0 . AA . A A on #h Feb. ..A A - 00 0 0 sepi. I I i lilli 80 120 160 100 A Fig. 6. Seatter diagrams plotting snout-vent length against month of col- lection: A, T. coronata; B, T. reticta relicta; C, T. reticta neilli. Solid symbols are mature individuals; hollow, immature; divided, apparently pubescent. Circles represent males, triangles, females. 1966 VARIATION AMONG SOUTHEASTERN SNAKES 298 Aug... 6 • • •• A O 0 •AA*OA A m on th m on th Feb. % . A. 0 66 0 0 A A t S.P'. 0 A 1111111 80 120 160 200 B Aug. A A I AI O %0a 0 Al A .A l A . Feb. A A ;AA. t 0 0 A A A O 006 6 00 A Aod • Sept. 09~ 0 .A Il lilli 80 120 160 200 snout-vent length. (mm) 294 BULLETIN FLORIDA STATE MUSEUM VoL 10 bama. Hardy (1952: 188) reported five specimens from Roundtop Mountain, Rutherford County, North Carolina, as "taken in a cleared field beneath rocks." Schwartz (1958: 156) collected the holotype of T. coronata mitri~er "beneath a Hat rock, about 18 inches in diameter" on a slope wooded with Pinus echinata and Ilex opaca. One of his paratypes was "caught on a wooded hillslope beneath a flat rock," two others "under one rock in a pine-oak woods,= and five were dug out of the ground from depths of 2 to 12 inches on a steep, rocky hillside. It seems rather clear from these accounts that while T. coronata may burrow as most small snakes probably do, it .is char- acteristically secretive and is usually associated with forests contain- ing pines. The only habitat data available in the literature on Tantilla oolit- ica are those of Duellman and Schwartz (1958: 306), who state: "In southern Florida Tantilla has been collected only on the eastern rim. and on Key Largo. Individuals have been found beneath rocks and boards in sandy soil in the pine woods, in hammocks, and in edifi- carian situations. The specimen from Key Largo was dislodged from a rotten stump. From their color notes they appear not to have confused T. relicta pamlica from coastal scrubs in southeastbrn Flor- ida with T. oolitica. A specimen not available to Duellman and Schwartz was collected beneath a log in a pasture southwest of Miami, and the holotype was found beneath a board on a vacant lot in Miami. Dennis R. Paulson (personal communication) states the holotype and two others were -collected under logs and trash in an empty: lot with much shrubby growth and a few Pinus elliotii. The area was un- doubtedly pine woods before clearing, with oolitic limestone over- lain by Some sand. The Martin Co. record is probably more charac- teristic of central Florida habitat-the specimen was under the dead, fanned out on the ground, leaves of a Serenoa repens partially buried under St. Lucie Fine sand, in typical Pinus clausa and Cerati- ola habita." The Martin County specimen referred to is a Tantilla relicta pamlica, and the passage quoted well describes the habitat difference between the two species in southeastern Florida. SIZE AND SEXUAL MATURITY Upon plotting snout-vent length and maturity against month of collection (figure 6), some differences between T. coronata, T . r . neilli, and T. r. relicta become apparent. T. coronata and T. oolitica are the largest species, and T. relicta the smallest. Within T. relicta, the nominate subspecies is intermediate, and neini the largest, with 1966 VARIATION AMONG SOUTHEASTERN SNAKES 295 T. relicta pamtica smallest in size, to judge from the small sample of the latter available. Table 10 presents size statistics of the main populations of southeastern Tantilla. TABLE 10. SNOUT-VENT LENGTH (IN ·MM~ OF SOUTHEASTERN TantiNG POPULATIONS Smallest Smallest Mature Largest Juvenile Male Female Male Female T, coronata 78 184 153 190 217 T. oolitica 92 147 165 171 246 T. reticta relicta 77 120 124 166 187 T. relicta neitti 80 142 184 190 194 T. felicta painlica 99 129 146 160 176 T. coronata apparently becomes mature at about 130 mm, T. r neilli and T, r. relicta at 120 mm. The samples of T. oolitica and T. r. pamlica are too small to permit such analysis. Males mature at slightly shorter lengths than do females. I discern four size groups in T. coronata: juveniles less than 105 mm; immature, young and young adult individuals from 110 to 150 mm; adults from 150 to 180 mm; and very large specimens over 180 mm. In both the T. relicta populations considered, only three size groups may exist: juveniles and young individuals below 120 mm in both; in neilli adults from 120 to 170 mm and very large specimens over 170 mm; and in the nominate subspecies adults from 120 to 155 mm and large individuals over 155 mm. It seems from these approximations that all three species become mature during their second growth season, but do not necessarily breed until the following year. Force (1985: 651) found that the Tantilla gracilis population in Oklahoma (where the active season is shorter than in the southeast) is composed of three age groups, that males mature at the same age as females but at a shorter length, and that maturity probably is reached in the third growth season at about 2% years of age. DISCUSSION HEAD PATTERNS. With respect to the populations of Tantilla relicta, an intriguing question recurred frequently during this study: What is the relation- ship of pattern to habitat in this ·species?· .The .banded populations, T. r. relicta and T. r. pamlica characteristically inhabit scrub, a plant 296 BULLETIN FLORIDA STATE MUSEUM Vol. 10 a b C t' e Fig. 7. a, Coastal Plain T. coronata (UF 12887, Coffee Co., Ala.); b, T. oolitica holotype (UF 17826, Dade Co., Fla.); c, T. relicta reacta holotype (UF 12421, Polk Co., Fla.); d, T. relicta neillijholotype, (UF 12406, Alachua Co., Fla.); 6, T. relicta pamlica holotype (UF 12480-2, Palm Beach Co., Fla.). 1966 VARIATION AMONG SOUTHEASTERN SNAKES 297 association that grows on white St. Lucie Fine sand, and which is a seral stage in dune succesion (Laessle, 1958). The nonbanded T. r. neilli is found in both sandhills and mesic hammock, but is more common in the former, and possibly occurs only in those mesic ham- mocks bordering sandhills areas. In areas of syntopy, relicta and tieilli seem completely segregated ecologically from each other by this habitat preference, yet in the southern ridge section of Florida where nemi does not occur, relicta is found :in both scrub and sandhills. I suggest as a purely speculative hypothesis that perhaps selec- tion acts against uniform head patterns on white sand substrates. T. relicta is a "submerged basker." It lies, in captivity at least, just beneath the sand, with the head at the surface. A banded head pat- tern would tend to break up head outline on the white sand and render a basking TantiUa less obvious to avian predators, to whom these diminutive snakes should pose no greater problem to ingest ventrals males females b 150 - b =9 -131- -41 6 ~ * -S ~6,466*t " '0 - 5.. 110 ' sukoudels meles females 70 b ie-* + fE+C + 4- d C g ,.. 77 N - Fig. 8. Variation in ventrals and subcaudals: a, T. coronata; b, T. oolitica; c, T..felicta relicta; d, T. reticta neilli, e, T. reticta pamlica, f, T. reticta from Sea- horse Key; g, series possibly intergradient between relicta and neilli. Vertical line is range; horizontal kne, mean; hollow bar, two standard· deviations of. the mean; solid bar, twice·the standard error. Sample size is indicated beneath each histogram. 298 BULLETIN FLORIDA STATE MUSEUM Vol. 10 than a large earthworm. The only known instance of avian preda- tion on Tantilla was one D. Jenni (personal communication) found in a cattle egret (Bubulcus ibis). I suspect that many of the birds in- habiting scrub, sandhills, and hammock prey upon Tantilla. Sub- merged basking is perhaps less dangerous for small snakes in sand- hills and hammock, both habitars with abundant vegetation on the ground. The scrub, however, a fire-climax association, typically has wide areas of open sand bearing little or no vegetation. The ability to bask with less likelihood of detection from above would thus make scrub more habitable for banded Tantilla, and reduce the likelihood of competition by a possibly more effective, blackheaded competitor. The predominantly nonbanded condition of disjunct populations of T. relicta in Charlotte County and on Seahohe Key is conceivably correlated with the overgrown, mature scrubs found there. Perhaps as the banded pattern becomes less advantageous a simple Mendelian dominant pattern reduces the frequency of banded phenotypes in these populations. This may also account for the presence of non- banded T. relicta in the mature scrubs east of the St. Johns River. INTERGRADATION. Most of the specimens treated above as possible intergrades be- tween T. r. relicta and T. r. neilli are not truly int6rmediate between neilli and relicta, but show some similarities to both. For instance, head pattern and ventrals seem closer to neilli, while subcaudals and tail length closely approximate relicta. Most of these specimens are from Putnam, Volusia, and Seminole counties, east and south of Lake George. Others are from scattered areas in Marion, Lake, Levy, Citrus, and Pinellas counties. I suspect that relicta and neilli do not intefgrade along a continuous front, but that some gene flow occurs through secondary intergradation, in the sense of Mayr (1968). These two long-isolated populations have probably come into con- tact since the closing of the Suwannee Straits and intergradation may be occurring only in those areas of longest contact, notably in the area around Lake George. A single specimen is available from Eau Callie, Brevard County, which may represent a similar hybrid be- · tween the central populations and T. r. pamlica. In characters of scutellation it resembles panilica, but its pattern is nonbanded. In- tensive collecting in Volusia County from the St. Johns to the At- lantic Ocean and south to Cape Canaveral will probably provide a less -speculative -interpretation of.these Tantilla populations. For the time being it is reasonable to consider them intergrades. 1966 VARIATION AMONG SOUTHEASTERN SNAKES 299 a 4 males femoles 21 . 8 -8- 9 0 9 + 3 6 11 7 -k v 7 .0 - ll 15 .. Fig. 9. Variation in proportionate tail length, expressed as per cent of total length. Symbols as in Fig. 8. PLEISTOCENE INFLUENCES UPON SPECIATION. Probably the most significant influence upon present distribution of the Floridian herpetofauna was that of fluctuating Pleistocene physiography. Correlations of ranges of such Florida endemics as Ophisaurus compressus (McConkey, 1954), Stilosonia extenuatum (Highton, 1956), Neoseps reynoldsi (Telford, 1959), and Eumeces egregius (McConkey, 1956; Mount, 1961) have been made with ma- rine terrace levels, and the present study shows that the same isolat- ing phenomena,probably contributed to divergence within the Florida Tantilia coronata complex. Pertinent geological factors in the history of the peninsula have been presented by Cooke (1939), Laessle (1958), Puri and Vernon 304 BULLETIN FLORIDA STATE MUSEUM Vol. 10 Highton, R. 1956. Systematics and variation of the endernic Florida snake genus Stilosoma. Bull. Florida State Mus., Biol. Sci., 1 (2): 78-96. Jan, G. 1862. Enumerazione sistematica delle specie d'ofidi del grupo Calamaridae. Archiv. Zool., Anat., Fisiol., 2 (1): 1-76. Jan, G., and F. Soidelli 1866. Iconographie g6n6rale des Ophidiens. Paris. Tome premier, livr. 15. (1961 reprint, J. Cramer, Weinheim). Laessle, A. M. 1958. The origin and successional relationship of sandhill vegetation and sand- pine scrub. Ecol. Monogr., 28: 861-887. Maerz, A., and M. R. Paul 1950. A dictionary of color. MeGraw-Hill, New York. 207 p. Mayr, E. 1968. Animal species and evolution. Belknap Press of Harvard Univ. Press, 797 p. McConkey, E. H. . 1954. A systematic study of the North American lizards of the genus Oph- isatirus. Amer. Midl. Nat., 51 (1): 183-171. McConkey, E. H. 1957. The subspecies of Eumeces egregilts, a lizard of the southeastern United , States. Bull. Florida State Mus., Biol. Sci., 2 (2): 18-28. Neill, W. T. 1951. Notes on the natural history of certain North American snakes. Publ. Res. Div., Ross Allen's Reptile Instit., 1 (5): 47-60. Neill, W. T., and J. M. Boyles 1957. The eggs of the crowned snake, Tantilla coronata. Herpetologica, 18: 77-78 Puri, H. S., and R. 0. Vernon 1960. Notes of the surficial geology of central peninsular Florida. IN: Late Cenozoic stratigraphy and sedimentation of central Florida, p. 24, Southeastern Geological Society Guidebook, Tallahassee. Schwartz, A. 1958. A new subspecies of crowned snake (Tantilla coronata) from the south- ern Appalachian mountains. Herpetologica, 9: 158-157. Smith, H. M. 1940. Descriptions of new lizards and snakes from Mexico and Guatemala. c Proc. Biol. Soc. Washington, 58: 55-64. Snedecor, G. W. 1946. Statistical methods. Iowa State College Press, Ames, 4th edit., 485 p. Telford, S. R., jr. 1959. A study of the sand skink, Neoseps reynoldsi Stejneger. Copeia, 1959 (2): 110-119. Wright, A. H., and A. A. Wright 1957. Handbook of snakes of the United States and Canada. Ithaca, C6m- stock Publ. Assoc. Vol. II, p. 565-1105. Contributions to the BULLETIN OF THE FLORIDA STATE MUSEUM may be in any field of biology. Manuscripts dealing With natural }histdry or systematic problems involving the southeastern Uaited States 6r the Caribbean area are solicited especially Manuscripts should be of medium length--50 to 200 pages, Examination tbr suitability is made by an Editorial Board. The BuLLETIN is distributed worldwide through institutional subscriptions and exchanges only. It is considered the responsibility 6f the author to distribute his paper t6 all interdsted indifiduals: To aid.in this, fifty copies are furnished the author without c6st. PREPARATION OF MANUSCRIPT Highly recommended as a guide is the v6Iume: Conference of Biblogical Editors, Committee on. Form and Style. 1960. Style manual for biological journals. Amer. Inst. Biol. Sci., Washington. 92 p. Manuscrifts sh6uld be typewritten with double spacing throughout, with ample margins, and on only one side of the paper. The author should keep a copy; the original and a carbon must be submitted. Tables and legends of figures should be typed on sheets separate from the text. Several legends or tables may be placed on a single sheet. Illustrations, including maps and photographs, should be referred t6 as 'Tgures." All illustrations are reduced to a maximum of 4-1/4 by 7-1/8 inches. Size scales, wherever they are netessary, should be incorporated into the figure. References to literature should conf6rm with the bibliographic style used in recent numbers of the BULLETIN. Spell 6ut in full the titles of non«English serials and places of publicati6n. Footnote material should be kept to a minimum. Iiowever, provide copy for a footnote detailing the title, affiliations, and address of the author (see recent num · bers of the. BULLETIN). Manuscripts muit be accompanied by a synopsis-a brief and factual summary (not a mere description) of the contents and conclusions, which. points out the presence of any new information and indicates its relevance. In it list all new organisms described and give their ranges; indicate all taxonomic changes pro- posed. The synopsis, wrjtten in full sentences, should be concise, but completely intelligible in itself without references to the paper, thereby enabling, the busy reader to decide more surely than he can from the title a16ne whether the paper merits his reading. The synopsis will be published with the paper. It does not replace the usual conclusions or summary sectibns. It may also serve as copy for the abstracting services. Manuscripts and all.editorial inatters should be addressed to: Managing ,Editor of the BULLETIN Florida State Museum Seagle Building Gainesville. Florida